TWI360984B - Method for receiving an optical ofdm signal and re - Google Patents

Description

1360984 VI. Description of the Invention: [Technical Field] The present invention relates to a receiving method and receiver for an optical multiplex system, and more particularly to an optical orthogonal frequency division multiplexing (Optica) In the 丨Orthogonal Frequency-division Multiplexing system, the receiving method and receiver of the received signal are compensated by the estimated time offset and Chromatic Dispersion. [Prior Art] Orthogonal Frequency-division Multiplexing (OFDM) is not a frequency division multiplexing (Frequency-division) using digital multi-carrier modulation (£^丨丨爪卜modulation) method. Multiplexing, FDM) system. A plurality of orthogonal subcarriers (〇rth〇g〇nal sub_carrier, also referred to as subbands) are used to transmit data. The data is cut into a plurality of parallel data streams or channels corresponding to respective subcarriers. Each subcarrier is modulated by a modulation technique with a lower symbol rate (such as quadrature amplitude modulation (QAM) or phase offset telemetry). (PSK 'Phase Shift Keying)) for data transmission. In this way, the total data transmission rate can be obtained in the same bandwidth compared to the traditional single carrier (sing|e-carrier) (_丨加(8). Qing Hui test "1A picture" The comparison with the spectrum distribution of the conventional direct transmission and the orthogonal frequency division multiplexing transmission is separately shown in Fig. 1B. The direct transmission is different from (4) the maximum difference from the transmission _ the distribution of the bandwidth, "1A" is displayed. The direct transmission wheel occupies ^60984. The frequency 〇 is subdivided into four equal divisions of equal width in a manner of orthogonal frequency division. Each sub-band fl (ie, the aforementioned sub-carriers) is orthogonal to each other. Then, the new spectrum distribution will be as shown in Figure 1B. The frequency response of the band can be roughly considered flat as long as the sub-bands are substantially equal to the parent sub-bands. That is to say, riding each sub-fresh only needs a single-coefficient equalizer to overcome the attenuation and phase distortion of each sub-channel. In addition, since the data rate of each +-channel τ is much lower than the original direct The data rate of the transmission, the operating clock of the equalizer naturally also decreases in equal proportions. The technology used in the field of wireless communication, the problem often encountered is the multi-path effect. Multi-path effect will derive time-spreading and inter-symbol interference (10) er_symb〇1 vitamins, other The frequency selectivity (freqUency_selective) channel is defined by P. This frequency selectivity problem is usually solved by adding a guard interval (GUard interval) in each OFDM symbol (or Symma, SymM). The meta-cycle, occupying the bandwidth of the stomach for transmitting data. When the orthogonal frequency division multiplexing technique is applied to the optical communication system, since the light passes through the same optical fiber, the multipath effect of the optical orthogonal frequency division and multiple n Not significant 'but there will be problems with inter-channel synchronization and inter-symbol interference like multi-path when receiving 汛嘁 at the receiving end due to Chromatic Dispersion. For synchronization in optical FDM systems The estimation method will be different from the previous estimation methods on copper wire or wireless OFDM transmission. Therefore, it is necessary to develop a new synchronization estimation method for the fiber channel characteristics to avoid Estimate the error. 5 1360984 Regarding the OFDM system, the synchronization estimation performed by the receiving end (that is, the estimation of Timing Offset) can be seen in the patent No. 7310302 of the US Approved Announcement on December 18, 2007. Method for estimating time and frequency offset in an OFDM system, and July 7th, 2007, in the US The patent "Timing offset compensation in orthogonal frequency division multiplexing systems" 〇 In addition, related papers are Minjian Zhao, Aiping Huang, Zhaoyang

Zhang and Miang Qiu's paper (please refer to μ Qiu Tradng L〇叩 for OFDM Symbol Timing, IEEE VTC52003, pp.2435-2439, vol. 4,

Oct. 2003), and papers by Baoguo Yang, Khaled Ben Letaief, Roger S. Cheng and ZhiganS Cao (please refer to Iing Recovery f0r OFDM transmissbMEEE I Select· Areas C(10)) n.,v〇i 】8, N〇 】], N〇v 2003) The conventional technique proposes a time offset estimation method for a 〇FDM system for wireless or copper wires, but because the characteristics of the fiber channel are not exactly the same as those of the wireless wheel or copper wire, The estimation method of synchronization in the optical OFDM system will be different from the previous estimation method. Since the accuracy of the symbol boundary estimation affects the performance of the DMFDM frequency domain equalizer or channel estimation, which affects the correctness of the demodulated transmission signal, that is, the synchronization estimation and channel estimation techniques are good or bad. Will directly affect the performance of the entire 〇fdm system. Therefore, a new synchronization estimation 1360984 method must be developed for the characteristics of the fiber channel to avoid estimation errors. SUMMARY OF THE INVENTION In view of the above-mentioned orthogonal frequency division multiplexing technology applied to optical fiber communication, the medium of the σ 彳 唬 不同 is different, and a time gate estimation method suitable for the optical 〇 FDM system needs to be developed at the receiving end, and the present invention provides A method of estimating the optical dispersion in H, and the method of dispersion. 'Shifting the present invention proposes a method for receiving an optical orthogonal frequency division multiplexing signal, which is suitable for a light two cross frequency rural read optical orthogonal frequency division multi-engine shot for the emitted light'. The receiving method includes: (10): The conversion signal is - digital signal; S52: estimate _ touch _ - symbol boundary; object: _ symbol boundary removes the digital signal - guard interval and becomes a signal; move: convert the power by = ^ 立叶The service is a plurality of frequency domain subcarriers Y(k), and each of the frequency domain sub-carriers corresponding to the same-frequency domain subcarrier includes a plurality of gates L; the carrier signals are located, and the carrier signals are located 1 pay, where k is greater than or equal to and less than the integer called i, the carrier battles the same - the (four) these guide days TMM i she estimates the bias ^ and the estimate of the boundary between the two sides of the gap. Just say S57: at least three of Dong {virtual number x (k) and at least three #+; ^ - symbol (four)?丨 载波 载波 载波 色 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ), machine carrier Y(kl), Y(k2) estimation—first phase rotation 1360984 slip; S572: two other guide carrier signals x(k3), x(k4) and corresponding ones a frequency domain subcarrier Y(k3), Y(k4) estimates a second phase rotation difference; and S574: estimating the time offset according to the first phase rotation difference and the second phase rotation difference; wherein kl, k2 , k3 and k4 are both greater than or equal to κ_Ν/2 and less than (N/2M integer, N is the number of points of the fast Fourier (FFTsize), lc2>kl, k4>k3, and k24d=k4 7.3. The second end of the second wood has a left-nose nose-one-color dispersion constant. Shuming proposed that the receiving end of the optical orthogonal frequency division multiplexing system is suitable for receiving a first orthogonal frequency division multiplexing system - the transmitting end - The optical signal, the receiving end package ^ electrical component, analog to digital component, the domain domain estimator, the guard interval vertical conversion component, the sigh _ reevaluation (four). Photoelectric conversion today = connection _ change _ number is _ _ _ digital component conversion symbol ir. 4 - touch符. The symbol estimates the distance according to the touch signal. The guard interval removes the component to form a signal according to the symbol interval. Fourier contains a plurality of symbol phenotypes, _, the f-domain subcarriers contain a data carrier signal (Data) and the guide y '' of the 5 hai symbol contains the plural _ 蝴 一 符 符 = = = = = = An integer that directs the carrier signal. The time offset estimator will correspond to and be less than (at least 1 of the carrier signal) and at least two of the "'s eccentricity. The symbol boundary _== the subcarrier m 测 测 测 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 By estimating the boundary of the symbol, the signal can be received accurately. The domain dispersion constant is used to set the length of the anti-sharp (G-w-) (when the guard interval is occupied (ie, the bandwidth), And improve the bandwidth used to transmit data. "About the features, implementations and work of the present invention, The detailed description of the preferred embodiment is as follows: The description of the contents of the above, the description of the above, and the description of the following implementations are based on the use of the non-fc and the transfer date, and further the scope of the patent application of Tixian (4). Interpretation. [Embodiment] Harmony test "2nd picture" is a schematic diagram of the receiving end architecture of the optical orthogonal frequency division multiplexing signal system of the present invention. The system of the present invention can be seen in the figure. It is suitable for an optical orthogonal frequency division multiplexing receiver 10. The receiver 10 receives the optical signal 98 transmitted by an optical orthogonal frequency division transmission benefit 90. The receiver includes an optical-to-electrical converting element (11), an analog to digitaj converting element, and a symbol boundary estimator (Symbol boundary estimator). A Guard interval removal element, a Fast Fourier Transform Element 丨5 (卩(10) Fom'ier Transferring element), and a time offset estimator 16(7) such as 哗offset estimator). 1360984 For the generation method of the optical signal 98 of the aforementioned transmitter 90, please refer to "3rd picture". It is a schematic diagram of the architecture of the transmitting end of the optical orthogonal frequency division multiplexing system according to the present invention. It can be seen that the transmitting end (i.e., the aforementioned transmitter 9A) includes a series of rotating and paralleling elements 91, a guiding carrier insertion component 92, an inverse fast Fourier transforming component 93, a guarding interval adding component 94, and turning together. The string element %, the digital to analog element 99, and an electro-optical conversion element 96 (Electrical-to-〇ptical converting element). The serial-to-parallel element 91 cuts the sequence digits to be transmitted by 匕π into a plurality of juxtaposed digital signals. Then convert each parallel digital signal into a parallel number of modulated by "Quadrature Amplitude Modulation" (QA.M 'Quarature A-tudeM〇duiati〇n) or "Phase Phase Shift Keying" (PSK PhaseShiftKeying). The digitized gas (four) and the component 91 are arranged in parallel with the modulated digital signal transmitted to the aforementioned button inserting element 92. The above-mentioned digital signals of ___ are all pending, and the data transmitted may also be referred to as "data carrier signal". The pilot carrier insertion component % then inserts/arranges the pilot carrier signal between the "data carrier L number" according to an appropriate arrangement, and combines the "data carrier signal" with the appropriately configured "guide carrier U tiger". The sub-carriers of the sub-carriers X_(k) belong to the sub-carriers in the frequency domain. That is to say, each subcarrier X(k) packet 3 has a plurality of "data carrier signals" and a plurality of "guide carrier signals". Off =, the configuration of "guide carrier signal" between the "data carrier signal" and the reference "4th picture" are read. Fig. 4 is a schematic diagram showing the configuration of the data carrier money and the pilot carrier signal in the subcarrier X(k).

The basic time length of the Vm p pilot carrier signal and the data carrier signal is 1360984 • The unit is defined as the symbol mbo or the code). This symbol is the minimum unit time length of the aforementioned sequence converted to analog signal. The "4th figure" order is the time axis, the unit is the symbol, Μ indicates. And it is recorded as frequency. It can be seen from Fig. that there are #16 circles in each vertical key column (c〇lumn). The horizontal axis corresponding to the dot on each vertical sharp column is called - one frequency domain subcarrier χ (_ in yeah), x(k2), x(k3) and x(k4)). Each subcarrier contains a plurality of symbols. The symbols corresponding to the same subcarrier x(k) include _ complex_data carrier signals 40a, 40b, and a plurality of pilot carrier signals 42a, 42b. The solid dots in the figure represent the pilot carrier signals 42a, 42b. The hollow dots are data carrier signals 40a, 40b. The pilot carrier signals 42a, 42b corresponding to the same symbol period (that is, the same symbol time length) are respectively distributed in each of the frequency domain subcarriers X(k). Seven identical symbol intervals represent the length of time per unit symbol on the horizontal axis. That is, the same queue (co[umn) is referred to herein as the same symbol interval. Therefore, "the pilot carrier signals 42a, 42b are respectively distributed in each of the subcarriers X(k) corresponding to the same symbol interval, that is, all the columns on the corresponding guide carrier signals 42a, 42b. The symbols are all pilot carrier signals 42a, 42b without any data carrier signals 40a, 40b. In addition, taking "苐4 diagram" as an example, there are three physical carrier signals between every two pilot carrier signals 42a, 42b on the same frequency domain subcarrier X(k) (ie, the same horizontal axis of the _ type). 40a, 40b. The present invention is not limited thereto, and the data carrier signals 40a, 40b between each of the two pilot carrier signals 5a 42a, 42b may be adjusted according to the system, and may be 1360984 but not limited to 1, 4 and 8. Then, the frequency domain subcarriers X(k) are converted into a plurality of time domain subcarriers Μ» by inverse fast Fourier transform. The plurality of time domain subcarriers <k) are described as having orthogonality with each other. Here, the size of the inverse fast Fourier transform (the number of points, (4) is equal to the number of sub-carriers in the frequency domain. Taking Figure 4 as an example, the inverse fast Fourier transform is 16 in size. Guard interval adding element) is to add a Cyclic prefix to each symbol month or to add a post-fix after each symbol. Parallel to Serial Converter, P/S) converts the shunt subcarrier x(k) with the guard interval into a (four) signal and transmits it to the digital to analog component 99. The digital ratio (the piece " the signal is converted to a finer ratio and then transmitted to the power The light-converting element 96. The electro-optical conversion element 90 can be, but is not limited to, a laser. The electro-optical conversion element will (4) collide with the optical signal 98 and transmit it from the ray. Refer to "2nd picture". The conversion component 丨] after receiving the optical signal 98, the paste _ _ 8G. The photoelectric conversion element 丨 can be, but not limited to, a light receiving gram (〇Pt) Cal receiver. The analog signal 80 is an electrical signal. Ride] 2 receives the aforementioned analog signal 8G and converts it into a radio number 8] Then, the complement compensator 13 estimates the position of the weaving position 峨81 to be transmitted to the Lai area by the wei wei _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The guard interval in Wei number 81 becomes a digital signal number 82. ^60984 Fast Fu Zeng (4) 15 Chi _ 峨 速 速 is a plurality of frequency domain subcarriers 83. Since the frequency received here is not transmitted continuously If the light is affected by the thief noise, etc., the frequency domain subcarrier X(k) of the _ _, transmitted _ has been turned over, so here, the received frequency domain subcarrier illusion vn _ system, miscellaneous. The service area subcarrier 83, time = test (four) is to receive the aforementioned frequency domain sub-cafe and estimate the time offset t where the time offset belongs to the remaining time miscellaneous esidua| = offset). That is, the symbol round 3 _ just mail error After: two = after, the time offset is transmitted to the symbol boundary estimator 13 to compensate the 兀 boundary estimator 13. The frequency domain subcarrier Y received by the month k) In addition to the domain subcarriers, it also contains (4) the frequency information transmitted by the homesickness. In the field of optical communication, due to the light 1 Response) and the miscellaneous ". There will be dispersion Μ. Dispersion in the domain channel cut into her bias, a ί璇俨, the dispersion constant of the work m, the phase offset generated for each pass is different. Therefore, the frequency domain subcarriers can be Expressed as: then domain receiving

Yi(k) a j,2· -jin- •Xi00 + PF\k) .(1) where 1 is the number of symbols (ie, the first symbol). N ^ number. » The point for the Fourier transform is the number of subcarriers (ie, k subcarriers).尺丁战和)κ is an integer of /2 1360984. W is the signal after the fast rifle in the first slap of the second squad. Chi is in the first! The multipath Channel Response of the kth subcarrier. B is the dispersion constant (d 丨 θ constant due to chr〇matic dispersion) 〇 r

"/(众).I is Ding on g 0). Training is the first! The transmission signal of the (10)th subcarrier of the character ^ may be the data carrier signal 'may also be the first channel of the carrier carrier'. The k-th subcarrier of the integrated channel is included in the multi-path of the dispersion disk. The resulting channel response, (4) is the ith message. The circuit block diagram of the time offset estimator 16 is shown in Fig. As can be seen in the figure, the time offset estimator 6 includes the first phase rotat_ est bribe t〇r, and the left estimate first phase difference estimating component 32 (e second phase r〇, tlon o . : Set extractor), and - dispersion constants (10) fine (10) theory). - 38 coffee. The paradox d sion time offset estimator 16 estimates the frequency domain subcarrier Y(k) when performing the time offset estimation. Estimate both —’ and U to. Since the (4) Mishan mother channel (subcarrier) is transmitted by the _ domain subcarrier signal signal sent by the hair (4), the button is the content of each pilot carrier signal. Therefore, the receiver is called (4) the carrier frequency domain subcarrier of the carrier bias. If the "1360984 4 picture" is taken as an example, the frequency domain subcarriers transmitted on the pilot carrier signal 42M2b_column (four) are transmitted. Male) is known for the receiving end (ie, receiver) Q). In the following description, the pilot carrier signals on the k4 subcarriers (bands) of the K1 subcarriers will be represented by X(4), which), X(4), and __. The above formula (1) towel. (4) is the 后 after the supply of the Fourier subcarriers in the Hgth symbol. Righteous to pure hardware (Th_ai

Effect) or environmental noise. It is less likely to change due to different symbols and different sub-channels, and the noise is not large. In the time offset estimation, it can be neglected for calculation convenience. The younger brother-accord element % is based on two of these guided carrier signals, X(k2) and the corresponding two frequency domain subcarriers, γ (clear calculation - first phase rotation difference angle (E(k2, kl)). The first-phase difference estimated element (4) will be returned to ==== to obtain the first-phase rotation difference:, :)). It: 尤(七2)” z(9) knife is the channel response of kl, k2. The estimation formula of the first phase rotation angle(E(k2, ld)) is as follows (2): angle{E{k.2,k \)) = angie{c〇,JΙΰ^ί) Hk\), _ ^ U^2)J .....................(2) The two phase difference estimating component 32 takes the other two pilot carrier signals X (four) and the _ field sub-key Y (k3), and takes the second phase rotation difference angle (E(lc4, k3)). Estimate the component 3 back ^ week ^..., yeah 4 generalized (8)) enter the hard number of the total vehicle domain and take its limbs 彳 _ two-phase rotary forging _ (_3)). Among them, ^I and population 丨 & J(/c4), χα, 3) The knives are the k3, k4 directional response. The second phase rotation difference angle(E(k4, k3)) is estimated as follows: angle{E^M)). angleicoJ2 is called) {^(kA)j X(k3) ..................(3) After the above -, cage _ " ¥ - _ Bu Le - difference Estimating the components 30, 32 gives the first-phase (four) value. The relationship between the two phase rotation differences and the time offset and dispersion constant can be obtained by the above-mentioned formula, and the ship is as follows: ' ^ The estimation of the slip is taken as an example, and the formula (1) is applied in the first symbol. The signal of the _ μ carrier after solving the Fourier Fourier (Eq. (4)), and the signal after the Fast Fourier is solved for the first and second subcarriers of the first symbol t (Equation (9) indicates /iw · Ί·k ΙίλΊ-»- —)β

J, 2.TT I:\-t ' e .(4) yM2) klik2+aui klr • e X , (^2)+ w,{kl) , medium, -N/2Sld<(N/2) -l. _j\j/;2gk2<(N/2)-l. As described in the section 'To simplify the calculation, the noise is ignored first. Multiply the 共(10) ί- ^ , 彻广尤(4)) 嗖 conjugate and take the angle of the knot as follows (6): .(6) ^ngle{E{klyk\)) s angle{coniΙ^Λ (ΣΜ U(々2)JU(H) where E(k2,ld) is the product of the complex number on the complex plane, and its calculation formula is as follows: 16 1360984 E{k2, Jc\)) two conj

(Y{k2) ^ (Y(k\) \ U(U)J Uw J

K2(k2+^~)B conj k\-( ) fj N ΛΓ =conj^ilO))-H]{k\) ^ Then calculate the angle of equation (7) to obtain the following equation (8)(10)(four) male 2, called) three g {Μ.r + 阳.(10)+令)_ /d Detect + contend] 々矣, Ak = k2-kl. In a similar manner to the derivation of equation (4), the relationship between the second phase rotation difference and the time offset 'dispersion constant can be obtained, as shown in the following equation (9):

N (7) (8) angle(E(k4,k3)) π n {A/cr + [/f4-(/c4 + -)-/c3.(/c3 + ^)].jg} .(9) where 'M = , fat _ (ie)) and shame __3)) are known. Therefore, 'by the calculus of equation (8), we can get the time offset r as the following formula (10): (4) which (four) 2 > Yu Lin (four) {[Μ. (Η + 全)-«—[々2:2 + order)—'[λ-4 · {kA + A) _ · (k3 + ^-)] · ~ angle{E{k2.. k\)) {[lc4-(k4 + ~)-k3.ik2 + ^)] - {k2 {k2 + y) - · (ΑΊ + y)]}^ [k2 (k2 + _) _ ^-] . + . . ailg[e^E(kA, lc3)) {[ ^4 · {kA + q2 + Z^.)] _ (lc.2 +-) - k] (k\ + —)]}. Ak 2 2 2 -—一~^ ___2. 2 · n .(10 Where 'kl,k2, k3 and k4 are both greater than or equal to _^/2 and less than (N/2)_i 17 ^〇U984 TM, and a difference is estimated to be 36 lang - _ and two phases of rotation = For the offset r. The time-deviation element 36 is about to be known as the first-, second-phase = turn _e (咏2, kl)) and the ton 四 (four) ^ offset value. ‘Human (丨G) can estimate the time of the closure of the 掏 掏 元件 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 帝 。 。 。 。 。 。 The graph device ι Γ Γ ° 36 includes - the - multiplier, - the second multiplication (_ first half of the value of the front of the formula). The first = 〇 is calculated as the formula (_ the second half of the value (that is, after the minus sign, the formula H method 8 362 is calculated - the multiplier 360 blood first - τ " brother - subtractor 364 is (4), such as the average (four) 丨 丨 丨 烂 上 上 ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( 啊 啊 啊 啊 啊 啊 啊 啊 啊 啊 啊 啊Detector]3 (such as "9th estimator" 3 in the next # _, _ ~ map"). This makes the boundary of the symbol and makes the Weiwei mining element = "Boundary touch, which is more correct, interval. More precise removal of the protection in the digital signal 81: the first dispersion, B < estimation" The dispersion constant _ calculation element 3" into -, according to the "phase rotation difference" and 4 pieces of 38 to enter the line. The above equation (8)_(9) is performed to estimate the dispersion constant B. After the same, -, the dispersion constant Β is the following formula (u) 18 丄 〇 984

N

B 2· π [angle{E{kA, kVj) - angle(E(k2, /cl))] [^4-(H + ^-)-/c3.(^3 + ^)]_[/c2 .(/c2 +y)-M(A:l + y)] N 2 π [angle(E(k4, /c3))] ^k4,(k4 + ~)-k3-(k3 + ~))- [k2-(k2 + ~)-k\(k]+^] __士[—导2,_ N-))-[k2-(k2 + ^)-k\.(k], (11) [" · (" +令)-/:3 . (3 + λ' = the angle in the equation (11) (E (k4, kiss.) The third multiplier is to reduce the aforementioned

[Material. The dispersion constant b can be obtained. For the circuit block diagram of the dispersion constant calculation element 38, please refer to "Figure 7". As can be seen in the figure, the dispersion constant calculation component 38 includes a second subtractor 38 and a second multiplier 382. The subtractor 380 subtracts the angle (E(k2, (10) multiplied by the dispersion constant B in the equation (n) to determine the length of the __. The multi-channel effect residual in the optical communication Instead, it is caused by the dispersion of the 屮^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The length of the code or the post = her code can be more appropriately determined. So, instead of the multi-channel effect problem caused by the benefit method, the length of the unnecessary guard interval is increased, thank you __ Marriage width. It is said that by dispersing 19 1360984, it is better to apply the bandwidth to transmit the data, and reduce the super _ although the inter-object offset estimator 6 contains the dispersion constant reckoning component%, but The % dispersion constant calculation element % may also be removed to include only the first estimation component 30, the second phase difference estimation component 32, and the scale calculation section 36. The object of the present invention may also be achieved. Just as kl, k2, k3, k4 have different names, but it does not show that the invention must be widely used with at least four Lai New Y (kl),), Y(k3), Qing) and four correspondences leave the remaining conditions of 1, 3 k2 ' unchanged. In this way, three frequency domain subcarriers (8)) 'Y(l, 2), Y(l.), Y(k4) and three corresponding pilot carrier signals 四(4), and (4) to 3) can be used. Χ(Η) can estimate the time offset and dispersion constant. In addition, the object of the present invention can be achieved by using more 'frequency domain subcarriers and pilot carrier signals. In view of the architecture of the receiving end of the optical orthogonal frequency division multiplexing system proposed by the present invention, the present invention further provides a method for receiving an optical orthogonal frequency division multiplexing signal according to the present invention. Yue Mai read "Figure 8". This method of acquisition is applicable to an optical orthogonal frequency division multiplexing reception. The receiver receives the optical signal transmitted by an optical orthogonal frequency division multiplexing transmitter. The receiving method includes: Step S50: converting the optical signal into a digital signal; Step S52: estimating a symbol boundary of the digital signal; Step S54: removing a guard interval of the digital signal according to the symbol boundary - digital telecommunication number; 20 1360984 step S56: convert the telecommunication number into a plurality of frequency domain subcarriers Y(k) by fast Fuli Rongshi _l tea million generation, and the number of sub-wallets of each domain sub-wallet is on the lang-frequency domain subcarrier The symbols include a plurality of data and a carrier signal, and the pilot carrier signals X(8) are located in the same symbol (4), and the tk is a large domain equal to -n/2 and less than (the integer 'N of N/2H is fast Fourier transform____ringring size); Step S57: corresponding to at least three of the pilot carrier signals X(k) and at least three corresponding frequency domain subcarriers corresponding to the same one Y(k) (the following is an estimate of a time offset r; and; step S58: using the time offset to compensate - 乂 + soil ", the estimate of the payout boundary, meaning that the check Then, step S52 "estimates the symbol boundary of the digital signal". Step S50 is performed by In Fig. 2, the output of the electric junction 11 and the analog digital transposition element 12 are completed. That is to say, the step (10) is further included (please refer to the "step S500: the optical signal is converted by photoelectric conversion". A type of analog signal S502 and a 5 Hz analog signal are converted into a digital signal. The analog step S500 is performed by the photoelectric conversion element J. The step _ is performed by the digital component 12. The estimated value of the step S52 is performed. The signal---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- The sequence is 丝 仃 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤 步骤_Measure-Time Offset 4 Time Offset Estimation After the value of the time offset is obtained, the value offset step of the time offset is as follows - this is the aforementioned step slave. That is, step S52 Estimating the digital signal by using the $0 boundary Compensating and estimating the symbol boundary of the digital signal. For the month, refer to "Figure 10". For step S57, the system includes: ^Step (5) 0: Guide the carrier signal X(kl) with two keys, X(k2) And corresponding frequency domain subcarriers Y(kl), 丫(4) estimate - first phase rotation difference; y'S: > 72. with the other two pilot carrier signals x(k3), χ(^) And corresponding to the frequency domain subcarriers Y(k3), Y(d) estimating - the second phase rotation difference; and S574 - estimating the time offset according to the first phase rotation difference and the second phase rotation difference. Where kl, k2, k3 and k4 are greater than or equal to small /2 and less than (n/2)-1 ^ ' is the Four Fourier Transferring size, k2>kl 'k4>k3 ' and Ic2-ld=k4-k3. (called i Y(kl) \ U(4)J U(10)j The first phase rotation difference of step S570 is the above equation (6): angle(E(k2, k))) s angle(conj The second phase rotation difference of step S572 is:

Γ Y(k2) (rm \ U(^2)J U(四)J an§ie(E(k2,kl)) ^ angle(conj 22 1360984 Inter-turnover is calculated according to the above formula (1) to formula (U) , can be estimated as: (9) + 矽]. order. -e (£ (9) called), 2, $_k, .«, al1gHE (kU3y)

{[A4 · (Η + _ /:3 · m + ^)] - [A2 · (/c2 + ^) _ /:|. (k] + ^)]}. M where angle(E(k2,kl )) is the first phase rotation difference, angk(E(k4, k3)) is the phase rotation difference ' conj is a complex number and is lightly multiplied, and n is the number of points of the fast Fourier transform: m = /c2_h = h_ /c3, Y(kl), Y(4), γ_γ(4) are pairs r: the frequency domain subcarriers of kl, k2, k3 and k4 in the interval, x(ki), x(i(2), () and X(k4) is the corresponding carrier signal corresponding to the symbol interval in ki, Θ and Μ. , ... 丨心骤 S57 "to correspond to the same - at least three of the symbolic intervals of the guide money _ At least three turns should be ___ turn) Estimated time is Xin Xi T" Another included. According to the first ^: g ^ ^ ^ (four) Zhao beta test and W two-phase compensation estimate - color government hanging number Β. The dispersion constant is: 5 = (male 2, called)] 4 + contend - Μ ??) Bu (four) 2 + f 卜 (four) + police)] ^ 帛 崎 $ $, the brother one phase rotation difference, N is the nkl. K3. [The number of points for idling the leaf, the method for receiving the crossover multiplex signal is further included: 7 ''Poly S59. Compensation with this time offset - estimate a carrier ·-:== 23 苓Figure 11 "This step S59 includes · Step S590: Converting the time offset to - phase offset; and step Na: _ mismatching the money domain subcarrier. After step SS92, the compensated frequency domain subcarriers have been compensated for the time _ bit rotation caused by each shift, that is, (4) sent to the "equalizer" for subsequent operations. The test was carried out in accordance with the receiver 10 of the present invention and the receiving t. Please refer to "Fig. 2", which is a schematic diagram of the daily estimation results of the test performed by the receiver of the optical orthogonal cut-to-clock frequency system of the present invention. This 7 is deliberately adding a known time offset at the receiving end. In this test case, a transmission rate of ^^B,r.(one sampllng clock , ^ . 〇 & tear (mother (four) shipped 1 〇Χΐ 0 samples Sa_e) is used, and each parent symbol interval is It is 12.8 ns. In this test example, it is also tested with optical communication systems with different color numbers. The dispersion values are G TS, _ melon, and the horizontal axis number of the zi dynasty ® is 0 = B = 0 Ts. For the discussion (9). No. T Ting s. And so on, number = 〇〇 B = 0.2. π θ Χ 录 录 社 物 物 物 物 物 物 物 物 物 物 物 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 The time offset value obtained can be clearly seen in the figure. Under different = number, the results of her estimation are all at the receiving end: the effect of taking the fine, the influence of the 咐-Cai, but the misrepresentation of the genus Small, code estimation result. It is quite accurate. 24 丄 3 〇 4 4 ^ 「 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第The number of estimates and fruit readings. This test article: and "12th picture" _. "13th picture" in the horizontal axis is the default dispersion constant of the 4th code #u ' and "12th picture", Paste is a constant linear dispersion method under this invention, the estimated result ^. ⑽ unit also sampling time. FIG towel bladder Each point represents the difference between both constant estimation result dispersion 1 with a predetermined dispersion value subtraction f). From the "° ° σ, it can be seen that the difference between the estimation result of the dispersion constant and the preset value of the test is between Ts and 8xl〇ts, that is, the difference value is very small, and the difference value comes from the news and the first The influence of the fiber channel response. It can be clearly seen from the above test results that the receiver 1〇 or the receiving method according to the present invention can estimate the time offset and the dispersion constant with a local accuracy. #差. The source of these errors may be other general effects, such as polarization mode dispersion (CQ, PQlarizatiQn Μ - coffee noise. But from the test.) The mouth is known, the secret small error belongs to The scope of the reception still meets the requirements of the aforementioned industry. The value is ambiguous. When the system increases the noise, the quality of the estimation is reduced, and the data can be collected and the data is caused by the noise. The error average eliminates the 'ton' quality, which means that the information of more than three carriers can be bribed. The invention is disclosed above in the preferred embodiment, but it is not limited to any buckle. The cypress artist is 'without departing from the invention The spirit and scope of the field can be used to make some 5 noon changes and retouching. Therefore, the scope of patent protection of the present invention shall be subject to the definition of the patent application parade attached to the Children's Book. 25 [Simple description of the drawing] The figure and the figure are a schematic diagram of the comparison of the spectrum distribution of the conventional direct transmission and the orthogonal frequency division multiplexing transmission. - The 2nd series is a schematic diagram of the receiving end architecture according to the optical orthogonal frequency division multiplexing. The figure is a schematic diagram of the architecture of the transmitting end of the optical orthogonal frequency division multiplexing system according to the present invention. Fig. 4 is a data carrier of the frequency subcarrier Deng of the transmitter according to the present invention; h tiger and ? FIG. 7 is a circuit block diagram of a time offset estimator according to the present invention. FIG. 7 is a block diagram of a circuit block according to the present invention. The dispersion constant is calculated [Figure. The circuit block diagram of the component is shown in Fig. 8. The schematic diagram of the method for receiving the multi-frequency (10) number of the multi-frequency (10) according to the present invention is shown in Fig. 9 is a schematic diagram of the received optical orthogonal step S50 according to the present invention. Schematic diagram of the process. The 10th is the flow chart of the step S57 of the Weihongjiao of the (4). The method of the tiger is the first embodiment of the process (4) of the receiving light positive step (10) according to the present invention. The method of the frequency signal is in accordance with the method of time offset estimation of the test performed by the receiver of the optical orthogonal frequency division multiplexing system according to the present invention. Fig. 13 is a view showing the estimation results of the dispersion constants tested by the receiver of the optical orthogonal frequency division multiplexing system according to the present invention. [Main component symbol description]

10 Receiver 11 Photoelectric conversion element 12 Analog-to-digital component 13 Symbol boundary estimator 14 Guard interval removal component 15 Fast Fourier transform component 16 Time offset estimation benefit 30 First phase difference estimation component 32 Second phase difference estimation Element 36 Time-shifting operation element 360 First multiplier 362 Second multiplier 364 First subtractor 38 Dispersion constant calculation element 380 Second subtractor 382 Third multiplier 40a, 40b Data carrier signal 27 1360984 42a, 42b Guide Carrier signal 80 analog signal 81 digital signal 82 electrical signal 83 frequency domain subcarrier 90 transmitter 91 serial and component 92 pilot carrier insertion component 93 inverse fast Fourier transform component 94 guard interval addition component 95 and repeat component 96 electro-optic conversion component 97 Serial digital signal 98 optical signal 99 digital to analog component f〇 bandwidth wide subband

28

Claims (1)

Seven, the scope of application for patents: L A red cross-counter financial work minus wire, _--optical orthogonal frequency division multiplexing connection, the reducer is received by --optical cross-frequency multi-wei transmission _ emitted optical signal The receiving method includes: converting the optical signal into a digital signal; estimating a symbol boundary of the digital signal; removing the digital signal from the boundary of the digital signal to become a digital signal; converting in a fast Fourier manner The digital signal is a plurality of frequency domain subcarriers, each of the frequency domain subcarriers includes a plurality of symbols, and the symbols corresponding to the same-frequency domain subcarriers comprise a plurality of data carrier signals and a pilot carrier signal. The pilot carrier signals are located in the same-symbol interval, where ^ is greater than or equal to and less than (N/2H integer, n is the Fast Fourier Transferring size); (d) at least three of the pilot carrier signals, though, to the 彡, three corresponding 姊, 纽 雄 )) side-time offset r; and the time offset to compensate for the estimate of the symbol boundary. 2. The method for receiving an optical orthogonal frequency division multiplexing signal according to claim 1, wherein the step of converting the optical signal into a digital signal comprises: photoelectrically converting the delta signal into an analog signal. And converting the analog signal to a digital signal. 29 If the method for receiving the orthogonal frequency division multiplexing signal of the request item m is as described in the request item, the step of measuring the symbol boundary of the digital signal is based on the time offset and estimating the digital signal. Symbol boundary. 4. The method for receiving an optical orthogonal frequency division multiplexing signal according to the invention, wherein the frequency domain subcarrier is a frequency domain subcarrier juxtaposed on a time axis. a method for receiving an optical orthogonal frequency division multiplexing signal as described in the following claim, wherein the at least three pilot carrier signals x(k) corresponding to the same-the symbol interval correspond to at least two The step of estimating a time offset r in the far-frequency domain subcarrier Y(k) includes: using the two pilot carrier signals x(kl), x(k2) and the corresponding two frequency domain subcarriers Y(ld) ), Y(lc2) estimates a first phase rotation difference; and the other two of the pilot carrier signals Χ(] θ), X(k4) and the corresponding frequency domain subcarriers Y(k3), Y( K4) estimating a second phase rotation difference; and estimating the time offset according to the first phase rotation difference and the second phase rotation difference; wherein 'Icl, k2, k3 and k4 are greater than or equal to -N/2 And the smaller integer 'N is the Fast Fourier Transferring size' 'k2>kl,k4>k3, and k2-kl=k4-k3. 6. The method for receiving an optical orthogonal frequency division multiplexing signal according to claim 5, wherein the first phase rotation difference is: 'F(H), Jm), angle(E(k2, kl)) = angle (coni{ where the second phase rotation difference is: 30 1360984 Y(k3) Y(k3) igle(E(kA,k3)) Ξ, /♦ ·[]'('") • U(4) where the time offset r is: [A4-(A4- -)-A3(A3 + -)].. NN Λ7 -2-? 2jj^s^e(£('c2,/c1)) - [/c2 (/c2 + —) - /d - (A1 + —)] —--ang!e(E(k4.1:3)) {[k4 ·(H + -)-^. (/;3 + ^}] _ μ2 (/;2 + }2_ Λ). (/c, + ^)]}. ΔΑ - 2 2 ; and a: where the angle(E(k2, kl)) is the first phase rotation difference, the ngle (E(k4, k3)) is the second phase rotation difference, the c〇nj is a complex number of vehicles multiplied 'the N is the number of points of the fast Fourier transform, the - /cl - L · ^! 4 Wc3, The Y(kl), the Y(k2), the Y(k3), and the y) each _pair Linhong_ in the Id, the k2, the k3, and the k4 frequency, the X(kl) The X(k2) and the X(k3) New Zealand X(k4) are respectively corresponding to the reference carrier signal in the k1, the k], the k3, and the U, and further include: (5) The method for receiving the optical orthogonal frequency division multi-wang signal, wherein the far-frequency domain sub-ship corresponding to the same as the C is the same as the 纟 — 间 间 间 间 间 间 间 ' ' ' ' Y(8) estimates a time offset r of B according to the first phase rotation difference scale second extinction (four) estimates a dispersion constant 8. =: the first orthogonal frequency division multiplexing signal (four) method of which 31 1360984 —— [angle(E(k4, /:3)) - angle(E(k2, ^1))] __λ ' 7t_____ [M · (k4 + —) ~ ^ · (^3 + γ)] - [^:2 · (k2 + -)- hi (^:1 + —)] - - 2 2 ^gle{E{k.2,k\)) = angle(conj ^ngle(E(kA,k3)) = angle( Conj (Y(k2) ^ fm) \ 1^2) J {xmj f Y{kA) λ f m3)] U'(")JU^3)/ where angle(E(k2,丨d)) is the First phase rotation difference, angle(E(k4,k3)) For the first phase rotation difference, N is the number of points of the fast Fourier, and the seam = k.2 ~ k\ = k4 ~ k3. 9. The method for receiving an optical orthogonal frequency division multiplexing signal according to claim 1, further comprising the step of compensating the frequency domain subcarriers with the time offset. The method for receiving an optical orthogonal frequency division multiplexing signal according to claim 9, wherein the step of compensating the frequency domain subcarriers by using an edge time offset comprises: converting a s-time time offset to a phase offset Shifting; and compensating the frequency domain subcarriers according to the phase offset. 11. A receiver for an optical orthogonal frequency division multiplexing system, adapted to receive an optical signal from a transmitter of an optical orthogonal frequency division multiplexing system, the receiver comprising: a photoelectric to H system receiving And converting the optical signal to - analog signal type analogy to digital device 'to convert the shame signal to - digital signal; - character 70 boundary estimate H, estimated by digital signal - symbol boundary; , according to the boundary of the element, remove the guard band of the digital signal 32 1360984 to form a telecommunication signal; : the idle Fourier transform component is converted by the fast fast transfer technique to reduce the _ subship Y(k)H , (4) (4) should be the same frequency domain sub-cut wave of the second two = in the same - symbol interval, its tk is greater than or equal to = , Fang Gang · (10) number 'N is the number of points for fast pass-through transformation, and a time offset estimator, which will correspond to the same _ the symbol interval of the symbol (four) wave X___ domain neutron γ (ι_ - The time-element boundary estimate is compensated and estimated according to the _time_shift. 12. The receiver described in U.S., the time offset estimator comprises: Estimating the component 'with two of these pilot carrier signals '(k2) and the corresponding two frequency domain subcarrier sides, Qing estimates a first phase rotation difference; - the second phase difference estimation component 'talks to the other two of the pilot carrier signals, and the corresponding The frequency domain subcarriers are purchased, and a second phase rotation difference is estimated; and the biasing material r 7G piece is estimated to be offset according to the first phase rotation difference and the second phase rotation difference; wherein砹匕 and ^ are integers greater than or equal to (10) and less than _-1, N is fast Fu Li #驰嫌, wrist, figure, 33 】 3 • as described in claim 12, iia 盥] plus). . ', έχ 相 相 估 估 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相 相The second phase rotation difference is the time offset component of the following formula: μ-4·(Α-4+ -)-^3-^3 '〇ngle(E(k2,k\)) ~[k2-(k2 + - k] (Λ) + ^}] N angle{E{kA,k3)) Where angle(E(k2,kl)) is the first phase rotation difference, angl<E(k4,k3)) is the far-phase rotation difference, N is the number of the fast Fourier points, M == 'Y( Kl), Y(k2), 丫_ and γ_ are respectively frequency domain subcarriers corresponding to kl, k2, k3 and k4·, x(k]) X(k2)s X(k3) And X(k4) are the pilot carrier signals corresponding to kl, k2, k3 and k4. 14. The receiver of claim 13, wherein the time offset estimator further comprises a Chromatic disPersi〇n constant extractor, according to the first phase 5 疋 slip and the second The phase rotation difference estimates a dispersion constant B. 34